基于重力数据的地质体埋深及物性三维反演方法研究

发布时间：2018-03-27 07:50

本文选题：自适应迭代　切入点：几何平均　出处：《吉林大学》2015年硕士论文

【摘要】：地质体埋深及物性参数的确定是重力数据解释的主要任务，是重力反演的前沿研究热点。本文针对确定地质体埋深的重力归一化总梯度法、DEXP-Depth（极点确定埋深）法以及物性参数的聚焦反演方法进行研究，以提高方法的实用性和结果的精度。重力归一化总梯度法无需任何先验信息情况下可以有效地获得地质体的空间分布。本文系统地分析了不同延拓和归一化方法对重力归一化总梯度反演结果的影响，通过模型实验对比分析，采用自适应迭代正则化方法和几何平均归一化方式的组合所得到的重力归一化总梯度反演结果具有更高的分辨率和精度。 DEXP-Depth反演成像方法能快速有效地估计地质体埋深，本文将其应用到重力梯度数据的地质体埋深计算，有效地压制背景场干扰，并将该方法扩展到重力三维反演计算。物性（密度）反演是根据重力数据来获得地质体的分布范围和属性特征，聚焦反演是现今较为实用和有效的方法之一。本文在重力异常数据聚焦反演中引入深度加权因子，，极大地减小了反演结果的趋肤效应，改善了聚焦反演的效果。在重加权正则化共轭梯度反演算法中引入深度加权和密度加权因子，通过模型实验和实际资料解释结果证明改进的算法反演结果可靠，具有较为广阔的应用前景。
[Abstract]:The determination of buried depth and physical parameters of geological bodies is the main task of gravity data interpretation. In this paper, the gravity normalized total gradient method (DEXP-Deptht) method for determining the buried depth of geological bodies and the focusing inversion method for physical parameters are studied in order to improve the practicability of the method and the accuracy of the results. The spatial distribution of geological bodies can be effectively obtained by gravity normalized total gradient method without any prior information. In this paper, the effects of different continuation and normalization methods on the inversion results of gravity normalized total gradient are systematically analyzed. Through comparison and analysis of model experiments, the results of gravity normalized total gradient inversion obtained by the combination of adaptive iterative regularization method and geometric average normalization method have higher resolution and accuracy. The DEXP-Depth inversion imaging method can estimate the buried depth of geological bodies quickly and effectively. In this paper, the method is applied to the calculation of the geological body depth of gravity gradient data, and the interference of the background field is suppressed effectively, and the method is extended to the three-dimensional inversion calculation of gravity. The physical property (density) inversion is based on gravity data to obtain the distribution range and attribute characteristics of geological bodies. Focusing inversion is one of the more practical and effective methods. In this paper, depth weighting factor is introduced into gravity anomaly data focusing inversion. The skin effect of inversion results is greatly reduced and the effect of focusing inversion is improved. The depth weighting and density weighting factors are introduced into the reweighted regularized conjugate gradient inversion algorithm. The results of model experiments and actual data interpretation show that the improved algorithm is reliable and has a broad application prospect.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.1

【参考文献】